pera

Nvidia's Cosmos world foundation models, post-trained on healthcare data (OpenH dataset), enable controllable synthetic data generation that grounds robotic policies in anatomical physics, with Cosmos Transfer, Predict, and Simulator creating a continuous learning loop that reduces development time from years to hours.

Nvidia is pioneering AI that processes raw RF signals (ultrasound) and k-space data (MRI) directly — bypassing traditional reconstruction — enabling real-time patient-adaptive imaging and potentially unlocking diagnostic information lost in conventional image-domain processing, analogous to CUDA's impact on reconstruction in 2017.

The same three-pillar foundation (world models, policy models, simulation) that enables surgical robotics is being applied to hospital automation via Project Rio blueprints, enabling digital twin assembly, data flywheels, and policy training for clinical cobots, supply chain robots, and assistive humanoids.

Nvidia's healthcare leader predicts 2026 as the year deployment challenges are solved and physical AI robots enter real clinical environments, driven by the convergence of world models, policy models, and high-fidelity simulation that closes the sim-to-real gap.

NVIDIA's healthcare lead states 2026 marks the transition from lab to deployment for surgical and hospital robots, driven by simulation-to-real pipelines, edge compute platforms, and partnerships with CMR Surgical, J&J MedTech, Moon Surgical and Peritas AI.

IGX Thor delivers 8x compute for future-proofing edge devices against VLMs/VLAs, while Holoscan 4.0 adds EtherCAT motor control, ROS 2 interoperability, and GPU-resident graphs for speed-of-light latency — creating a deployment stack native to physical AI workloads rather than legacy CNN pipelines.

NVIDIA introduces 'raw to insights' models that process raw RF/ultrasound/MRI signals directly with physics-grounded AI, potentially unlocking diagnostic information lost in conventional reconstruction pipelines, with production-ready ultrasound model built with Siemens and MRI model topping CMRx recon challenge.

Cosmos-based surgical simulator achieves policy ranking correlation with real-world performance, reducing simulator development cost and time by an order of magnitude versus classical physics simulation, enabling hybrid classical-AI simulation approach for surgical sciences.